Antibiotic resistance emerges when a mutation in a bacterium causes a previously inhibitory concentration of a compound to become survivable. Through the accumulation of mutations conferring varying increases in resistance, already many easy-to-treat infections have become nearly incurable, and are spreading in part anthropogenically. The classical model of resistance evolution, that a resistant mutant has a fitness advantage in the presence of antibiotic use, and so spreads in the population to near-fixation, captures the rise of antibiotic resistance, but on closer inspection fails to explain several critical features of resistance. First, antibiotic resistance rarely reaches fixation in clinical populations; more importantly, sensitivity is higher than the population-genetic models would predict. Second, antibiotic resistance was present, and likely common, in clinical infections before the human use of antibiotics even began. Third, despite the widespread prevalence of antibiotic-producing bacteria in the environment, these same bacteria remain surrounded by sensitive neighbors. For these reasons, we hypothesize that the existing model of resistance evolution is incomplete, and in particular that there exist evolutionary factors in the environment which have a potentially countervailing effect on resistance evolution of similar or greater magnitude to the human use of antibiotics. Here, we will combine evolution experiments in model systems with computational modeling and database mining of sequence data to study the constraints on the evolution of resistance, focusing on two key areas: the role of spatial structure in the evolution of resistance, and the role of selfish genetic elements including phages and parasitic plasmids. Resistance provides an almost ideal model system for the study of microbial evolution; fitness can be well defined, imposed selective pressures can be readily tuned, and can emerge either spontaneously or by horizontal gene transfer. We expect to uncover the evolutionary mechanisms behind the emergence, spread, and limitation of antibiotic resistance.